Test Partners: A Formula For Success

In reading a recent Innovation Abstracts, an article
entitled “The Algebra Cup” (Volume XII, No. 14), I was
a bit overwhelmed by some of the strategies and
activities that the author had undertaken in her College
Algebra classes. But the idea of the “Partnership
Exam” caught my attention. I had been looking for a
way to help my students overcome their test/math-
ematics anxiety and reduce their careless errors on
tests.

In the 1990 spring session, I allowed students in my
General Education Mathematics classes to take one test
in groups of their choosing. The results were mixed.
The material for the test was especially difficult, but the
test results were better than they had been in previous
years. I attributed the improved test performance to
the group setting. The students liked the group test,
but I had a problem with it. Students who had per-
formed poorly on previous tests chose groups with
stronger mathematics studenis and reaped the benefits.
I felt this was unfair to the weak student whose test
grade was inflated and the strong student who was
being “used” by the weak student. So, I discontinued
the process for the balance of the session.

After some thought and discussion with colleagues,
I decided to attempt the process again in all of my
classes (General Education Mathematics, Intermediate
Algebra, and College Algebra) with some modifica-
tions. This time students would be grouped with
someone performing at or near the same level. To make
that determination, students took the first two unit
tests individually. I then averaged the two test grades
and ranked them from highest to lowest (via a com-
puter spreadsheet). Next, guidelines for group testing
were outlined with each class: Students would be
assigned in pairs or triads with individuals who had
performed at the same level on the first two tests. Test
partners would take the test together and turn in one
set of results; each student in the group would receive
the same grade. Participation was voluntary.

From among the students wishing to participate,
students were assigned to groups. Groups generally
consisted of pairs of students of the same sex. (I
thought students might feel more comfortable working
with someone of the same sex; however, there were a
few groups with both males and females that seemed
to work well.) There were also some groups of three
students. The triads were formed either because the
spread of averages on the first two tests required it or
to minimize the impact of a student who had demon-

strated by personality, behavior, or performance that
he/she might not be an asset to a group.

No changes were made in the tests except for the
grouping of the students. The results were phenomenal.
In one Intermediate Algebra class, all the students who
had “test partners” scored as high or higher than they
had on previous tests. This is at a point in the course
when test scores generally start falling as the material
gets more difficult. The group with the highest previ-
ous averages achieved the maximum possible score, an
improvement of less than 10 points. However, stu-
dents in the other groups showed improved test scores
ranging from 10 to 40 points. One group of two
students who had consistently scored in the 50’s on the
previous two individualized tests showed the greatest
improvement (40 points). .

What accounts for these phenomenal results? I think the
answer is threefold. First, the test partner format
minimizes or eliminates test or mathematics anxiety.
Students have a support system if they get nervous or
“go blank.” Second, the group format reduces the
number of careless errors. I suggest that each person in
the group work a problem and then compare the
results to determine what they all believe is the correct
answer. Students comment frequently that they find
each other's careless errors. Hopefully, experience
with locating the careless errors of a test partner will
enable a student to find more of his/her own. Third,
the peer pressure of being in a group forces students to
study so they can “carry their weight” on the test. A
person who does not contribute meaningfully on test
day is likely to be abandoned by his/her test partner(s)
the next time a test is scheduled.

I feel comfortable with a test partner system in
which grouping is based on level of performance—
students’ test performances are not artificially inflated.
Group testing removes the obstacles of anxiety, careless
errors, and lack of motivation to learning and success-
ful test performance that many students experience.
With these obstacles minimized or eliminated, students
can truly perform at their best.

Theresa Geiger, Instructor, Mathematics
For further information, contact the author at St.

Petersburg Junior College, 6605 5th Avenue North, St.
Petersburg, FL 33733.

 

Suanne D. Roueche, Editor

April 5, 1991, Vol. XIll, No. 11

©The University of Texas at Austin, 1991
Further duplication is permitted by MEMBER
institutions for their own personnel.

INNOVATION ABSTRACTS is a publication of the National Institute for Staff and Organizational Development (NISOD),
Department of Educational Administration, College of Education, EDB 348, The University of Texas at Austin, Austin, Texas
78712, (512) 471-7545. Funding in part by the W. K. Kellogg Foundation and the Sid W. Richardson Foundation. Issued weekly
when classes are in session during fall and spring terms. ISSN 0199-106X.

ais SEN...

o